Skip to content

Provide a fast (cpp-version) of Prioritized Experience Replay in Reinforcement Learning

License

Notifications You must be signed in to change notification settings

7starsea/Prioritized-Experience-Replay

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Shark-PrioritizedExperienceReplay

MIT licensed

Inspired by https://github.com/takoika/PrioritizedExperienceReplay, we provide a fast (cpp-version) implementation of prioritized experience replay buffer.

for a full usage, please take a forward step to https://github.com/7starsea/shark.

Note

Implementation here is simplified version of PrioritizedExperienceReplay in project https://github.com/7starsea/shark by removing unnecessary bundle.

Requirement

  • python3, (tested on 3.6, 3.7)
  • pybind11, (tested on 2.4.3)
  • cmake3, (tested on 3.14)
  • scikit-build, (tested on 0.10.0), optional

(we use ananconda enviroment for testing)

Compile [tested on ubuntu]

on linux, you can type ./complie.sh or python setup.py build --inplace.

Reference

[1] "Prioritized Experienced Replay", http://arxiv.org/abs/1511.05952

About

Provide a fast (cpp-version) of Prioritized Experience Replay in Reinforcement Learning

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published